Constrained Optimization MCP Server

""" Tests for HiGHS linear and mixed-integer programming solver. """ import pytest import numpy as np from returns.result import Success, Failure from constrained_opt_mcp.models.highs_models import ( HiGHSProblem, HiGHSProblemSpec, HiGHSVariable, HiGHSConstraints, HiGHSObjective, HiGHSOptions, HiGHSSense, HiGHSVariableType, HiGHSConstraintSense ) from constrained_opt_mcp.solvers.highs_solver import solve_highs_problem class TestHiGHSSolver: """Test cases for HiGHS solver functionality.""" def test_simple_linear_programming(self): """Test solving simple linear programming problems.""" variables = [ HiGHSVariable(name="x1", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), HiGHSVariable(name="x2", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), ] objective = HiGHSObjective(linear=[1.0, 2.0]) constraints = HiGHSConstraints( dense=[[1, 1], [1, -1]], sense=[HiGHSConstraintSense.LESS_EQUAL, HiGHSConstraintSense.GREATER_EQUAL], rhs=[5.0, 1.0] ) problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MINIMIZE, objective=objective, variables=variables, constraints=constraints ) problem = HiGHSProblem(problem=problem_spec) result = solve_highs_problem(problem) assert isinstance(result, Success) solution = result.unwrap() assert solution.status.value == "optimal" assert solution.objective_value is not None assert "x1" in solution.values assert "x2" in solution.values def test_mixed_integer_programming(self): """Test solving mixed-integer programming problems.""" variables = [ HiGHSVariable(name="x1", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), HiGHSVariable(name="x2", type=HiGHSVariableType.INTEGER, lb=0, ub=10), ] objective = HiGHSObjective(linear=[1.0, 2.0]) constraints = HiGHSConstraints( dense=[[1, 1], [1, -1]], sense=[HiGHSConstraintSense.LESS_EQUAL, HiGHSConstraintSense.GREATER_EQUAL], rhs=[5.0, 1.0] ) problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MINIMIZE, objective=objective, variables=variables, constraints=constraints ) problem = HiGHSProblem(problem=problem_spec) result = solve_highs_problem(problem) assert isinstance(result, Success) solution = result.unwrap() assert solution.status.value == "optimal" # Check that x2 is integer assert solution.values["x2"] == int(solution.values["x2"]) def test_binary_variables(self): """Test solving problems with binary variables.""" variables = [ HiGHSVariable(name="x1", type=HiGHSVariableType.BINARY), HiGHSVariable(name="x2", type=HiGHSVariableType.BINARY), ] objective = HiGHSObjective(linear=[1.0, 2.0]) constraints = HiGHSConstraints( dense=[[1, 1]], sense=[HiGHSConstraintSense.LESS_EQUAL], rhs=[1.0] ) problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MAXIMIZE, objective=objective, variables=variables, constraints=constraints ) problem = HiGHSProblem(problem=problem_spec) result = solve_highs_problem(problem) assert isinstance(result, Success) solution = result.unwrap() assert solution.status.value == "optimal" # Check that variables are binary assert solution.values["x1"] in [0, 1] assert solution.values["x2"] in [0, 1] def test_maximization_problem(self): """Test solving maximization problems.""" variables = [ HiGHSVariable(name="x1", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), HiGHSVariable(name="x2", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), ] objective = HiGHSObjective(linear=[3.0, 2.0]) constraints = HiGHSConstraints( dense=[[1, 1], [2, 1]], sense=[HiGHSConstraintSense.LESS_EQUAL, HiGHSConstraintSense.LESS_EQUAL], rhs=[4.0, 6.0] ) problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MAXIMIZE, objective=objective, variables=variables, constraints=constraints ) problem = HiGHSProblem(problem=problem_spec) result = solve_highs_problem(problem) assert isinstance(result, Success) solution = result.unwrap() assert solution.status.value == "optimal" assert solution.objective_value is not None def test_infeasible_problem(self): """Test handling of infeasible problems.""" variables = [ HiGHSVariable(name="x1", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), HiGHSVariable(name="x2", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), ] objective = HiGHSObjective(linear=[1.0, 1.0]) constraints = HiGHSConstraints( dense=[[1, 1], [-1, -1]], sense=[HiGHSConstraintSense.LESS_EQUAL, HiGHSConstraintSense.LESS_EQUAL], rhs=[1.0, -2.0] # This creates an infeasible problem ) problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MINIMIZE, objective=objective, variables=variables, constraints=constraints ) problem = HiGHSProblem(problem=problem_spec) result = solve_highs_problem(problem) assert isinstance(result, Success) solution = result.unwrap() assert solution.status.value == "infeasible" def test_unbounded_problem(self): """Test handling of unbounded problems.""" variables = [ HiGHSVariable(name="x1", type=HiGHSVariableType.CONTINUOUS, lb=0), HiGHSVariable(name="x2", type=HiGHSVariableType.CONTINUOUS, lb=0), ] objective = HiGHSObjective(linear=[1.0, 1.0]) constraints = HiGHSConstraints( dense=[[1, -1]], sense=[HiGHSConstraintSense.LESS_EQUAL], rhs=[0.0] ) problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MINIMIZE, objective=objective, variables=variables, constraints=constraints ) problem = HiGHSProblem(problem=problem_spec) result = solve_highs_problem(problem) assert isinstance(result, Success) solution = result.unwrap() assert solution.status.value == "unbounded" def test_solver_options(self): """Test solver with custom options.""" variables = [ HiGHSVariable(name="x1", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), ] objective = HiGHSObjective(linear=[1.0]) constraints = HiGHSConstraints( dense=[[1]], sense=[HiGHSConstraintSense.LESS_EQUAL], rhs=[5.0] ) problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MINIMIZE, objective=objective, variables=variables, constraints=constraints ) options = HiGHSOptions( time_limit=10.0, output_flag=False, presolve="on" ) problem = HiGHSProblem(problem=problem_spec, options=options) result = solve_highs_problem(problem) assert isinstance(result, Success) solution = result.unwrap() assert solution.status.value == "optimal" def test_sparse_constraint_matrix(self): """Test solving with sparse constraint matrix.""" variables = [ HiGHSVariable(name="x1", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), HiGHSVariable(name="x2", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), ] objective = HiGHSObjective(linear=[1.0, 2.0]) # Create sparse matrix from constrained_opt_mcp.models.highs_models import HiGHSSparseMatrix sparse_matrix = HiGHSSparseMatrix( rows=[0, 1], cols=[0, 1], values=[1.0, 1.0], shape=(2, 2) ) constraints = HiGHSConstraints( dense=None, sparse=sparse_matrix, sense=[HiGHSConstraintSense.LESS_EQUAL, HiGHSConstraintSense.LESS_EQUAL], rhs=[5.0, 3.0] ) problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MINIMIZE, objective=objective, variables=variables, constraints=constraints ) problem = HiGHSProblem(problem=problem_spec) result = solve_highs_problem(problem) assert isinstance(result, Success) solution = result.unwrap() assert solution.status.value == "optimal" def test_problem_validation(self): """Test problem validation.""" # Empty variables problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MINIMIZE, objective=HiGHSObjective(linear=[]), variables=[], constraints=HiGHSConstraints(dense=[], sense=[], rhs=[]) ) problem = HiGHSProblem(problem=problem_spec) assert not problem.validate() # Mismatched objective and variable dimensions variables = [HiGHSVariable(name="x1", type=HiGHSVariableType.CONTINUOUS)] objective = HiGHSObjective(linear=[1.0, 2.0]) # 2 coefficients for 1 variable constraints = HiGHSConstraints(dense=[], sense=[], rhs=[]) problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MINIMIZE, objective=objective, variables=variables, constraints=constraints ) problem = HiGHSProblem(problem=problem_spec) # This should fail during solving, not validation def test_dual_values(self): """Test extraction of dual values.""" variables = [ HiGHSVariable(name="x1", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), HiGHSVariable(name="x2", type=HiGHSVariableType.CONTINUOUS, lb=0, ub=10), ] objective = HiGHSObjective(linear=[1.0, 2.0]) constraints = HiGHSConstraints( dense=[[1, 1], [1, -1]], sense=[HiGHSConstraintSense.LESS_EQUAL, HiGHSConstraintSense.GREATER_EQUAL], rhs=[5.0, 1.0] ) problem_spec = HiGHSProblemSpec( sense=HiGHSSense.MINIMIZE, objective=objective, variables=variables, constraints=constraints ) problem = HiGHSProblem(problem=problem_spec) result = solve_highs_problem(problem) assert isinstance(result, Success) solution = result.unwrap() assert solution.status.value == "optimal" # Dual values should be available for linear programming if solution.dual_values: assert len(solution.dual_values) == len(constraints.sense)